The Alkali-Agregate Reaction Problem
(Or, "Shucks, I wish I knew that before!")
When searching for information about some dams, we found that many of the hits we would get would be about consulting firms assisting dam owners with problems resulting from "AAR", aka Alkali-Agregate Reactions. We came to learn that this is a serious problem with some dams.
AAR results when the alkali contained in cement reacts with certain types of agregate, or fill, when the two are combined to make concrete. Only certain types of agregate are reactive which is why not all dams have a problem. One source lists the requirements for AAR to occur: a reactive agregate, high alkali content, and moisture. Dams will always have plenty of moisture! One description of the process states that the reaction produces a gell that absorbs moisture. The end result is that the concrete expands, or "grows". This places stress on certain parts of the structures that, if left unchecked, could ultimately threaten the integrity of the entire structure. Not a good situation for a dam to be in!
So, what do you do about it? From what we have read, most of the remedial action includes the cutting of relief slots periodically, and anchoring portions of the structure with specialized hardware. But most important is the constant monitoring of specialized instrumentation installed to detect the expansion and displacement of different parts of the structure. Of course, now that we know about the problem, you could tear the thing down and rebuild it with a non-reactive agregate... yeah, right!
Actually, there is one situation we learned of where that might happen. It seems that the entire Chickamauga project, TVA's dam and lock in Chattanooga, TN, has a severe AAR problem. One of the effects is the navigation lock has the most down time in the system, due to unscheduled maintenance required when problems are discovered. The Corps of Engineers and TVA studies indicate that they can safely and economically operate the lock until about 2010. At that time it will no longer be cost effective to repair, so the cheap route would be to plug it, and end commercial shipping to any degree upstream from that point. A Corps study looked at the economics of several different scenarios, including replacement. See a summary of the report here... this is a pdf file, and requires that you have a pdf reader installed. Long story short: the recommendation is to replace the lock.
Another interesting case is Santeetlah Dam, one of the dams in Alcoa's Tapoco Project. The main dam stared life as a concrete arch dam. Later, two concrete gravity buttresses were constructed, forming gated spillways at each end. Over the years, measurements seem to indicate that the dam was moving upstream! Turns out that AAR and seasonable temperature related expansion and contraction caused a ratcheting action between the central arch portion and the two butresses that was indeed moving the dam upstream. Naturally at some point a serious failure would have occurred, but the constant monitoring and remedial action keeps the dam safe and operable. Maybe not as economically as hoped, but cheaper than rebuilding it... which may or may not be permitted in these litigous times!
Hopefully, now that the problem is well known, the agregate will be chosen a little more carefully. In the past, many times the agregate was supplied from local quarries, as is very visible at Fontana, which explains why some have the problem, some don't. It all depended on the luck of the (local) draw.